Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 91
4 Existing Measures of Child and Adolescent Health Health is not bought with a chemist’s pills, nor saved by the surgeon’s knife. Health is not only the absence of ills, but the fight for the fullness of life. —P. Hein Prologue at the celebration of the 40th anniversary of the World Health Organization (1988), Copenhagen (Reprinted with permission by WHO) Summary of Key Findings • Multiple data systems capture information on specific health conditions, but there appears to be overlap in their popula- tions and content. Moreover, measures are inconsistent across states, and no current mandate exists for comparability and standardization. • Current data collection systems for monitoring health fre- quently fail to address important social and environmental factors that influence children’s health outcomes. Likewise, data collection systems that monitor educational performance or children’s well-being frequently omit health data. • Multiple recommendations for improving health measures for children and adolescents have emerged in recent years. How- ever, current federal surveys do not yet include a robust set of measures of positive health, functioning, development, and health potential within a life-course framework. • Significant disparities in health status and health care quality currently exist for a variety of racial, ethnic, and sociodemo- graphic populations of children. 91
OCR for page 92
92 CHILD AND ADOLESCENT HEALTH • Social and economic conditions influence child health. Such conditions include not only household income and educational level, but also such factors as racial and ethnic identity, family structure, immigrant status, urban/rural location, and health literacy. • Multiple environmental factors influence child health, many of which are outside the purview of the health care system. • Data on community factors are frequently available in non- health surveys (e.g., environmental surveys, educational sur- veys, or child victimization surveys). • A life-course approach provides a basis for understanding the relationships among early health conditions, health influences, and later health status. • Child health is strongly influenced by family and especially maternal health (e.g., maternal depression). The development of conceptually sound and reliable health measures for children and adolescents is of critical importance for policy makers, researchers, clinicians, and families, as well as community leaders and the general public. Child and adolescent health measures can be used to assess the effects of disease or injury on health; to identify vulnerable children in clinical practices and vulnerable population subgroups in health plans or geographic regions; to measure the effects of medical care, policy, and social programs; and to set targets for improving health care (Szilagyi and Schor, 1998). Health measures also can identify general health trends over time to highlight areas of progress as well as emerging areas of concern. Until the middle of the 20th century, data on infant and child mortality provided a reasonable assessment of child health (Guyer et al., 2000). The neonatal segment of infant mortality (number of infant deaths at less than 28 days per 1,000 live births) provided a window on conditions related to fetal development, complications of pregnancy and delivery, and the new- born period; the postneonatal segment helped in understanding conditions influencing child health through the preschool years (Black et al., 2003; Heron et al., 2010). The middle of the 20th century saw a decrease in the influence of infec- tious diseases on child health. A different pattern of morbidity emerged, termed the “new morbidity” (Haggerty et al., 1993; Palfrey, 2006). The conditions dominating child health today often reflect behavioral and de-
OCR for page 93
93 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH velopmental problems and chronic conditions, as well as associated social conditions, which are poorly captured in vital statistics systems. This same period saw the emergence of a wealth of measurement tools in developmental psychology for assessing normal child development, including Ages and Stages Questionnaires (ASQ), Bayley Infant Neurode- velopmental Screens (BINS), Parents’ Evaluations of Developmental Status (PEDS), and the Wechsler Preschool and Primary Scale of Intelligence (WPPSI), among others. The application of these measures, however, has been limited by both conceptual and practical issues. The conceptual issue is that theories of developmental psychology are still evolving and do not agree on the selection of appropriate domains for assessment. A comparison of several well-established child health measures, for example, reveals 14 separate dimensions of child health (Landraf et al., 1996). Moreover, many of the dimensions, such as learning disabilities, require sophisticated testing by trained examiners. Practical issues include provider time, reimburse- ment, and differential skill requirements for administering the instruments. Early efforts focused specifically on measures of child health status that would capture issues related to functional abilities were patterned after more well-established adult measures (Eisen, 1980; Starfield et al., 1993). For example, many adult health function measures inquire about the impact of health issues on work and can be adapted to inquire about school for older children. For preschool children and infants, however, such adapta- tion is limited, as the activities of younger children are focused more on attaining developmental skills necessary to attend school and participate in other activities. Further, data on the validity and reliability of even es- tablished measures are relatively sparse for pediatric outcomes. Validity is established most commonly by the ability of the instrument to yield differ- ent scores when administered to healthy children and those with established diagnoses. Most instruments have not been used in a longitudinal fashion, moreover, so that information on predictive validity is lacking, and little has been done to validate responses against clinical observations. For example, if a mother reports that her child has difficulty in play activities, does this indicate a lack of stamina, a lack of coordination, or a lack of social skills? Alternatively, does it reflect the mother’s lack of understanding of what developmentally appropriate play looks like at that age? Since the adoption of quality improvement initiatives under the Chil- dren’s Health Insurance Program Reauthorization Act (CHIPRA), as well as new quality efforts authorized under the Patient Protection and Afford- able Care Act (ACA), the Congress and public and private health agencies have begun searching for valid, reliable, and accessible health and health care measures that can support the implementation and evaluation of these efforts. Ideally, such indicators would provide the capacity at the national, state, and local levels both to monitor the overall health of children and
OCR for page 94
94 CHILD AND ADOLESCENT HEALTH adolescents and to analyze the quality of health care services offered to both the general population and vulnerable groups of children and adolescents. An ideal set of health measures would inform comparisons of the status of children and adolescents served by different health plans (both public and private) and the types of health issues associated with different provid- ers (pediatricians versus nurse practitioners and primary versus specialty care) and health settings (such as hospitals or ambulatory care settings). These measures would provide opportunities for states or regions of the country to monitor the conditions of children and adolescents in areas relevant to their own circumstances. Ideally, robust health indicators would reveal significant trends and changes in health status over time for the general population of children and adolescents, as well as special groups that are at particular risk for poor health outcomes and frequently are not identifiable in the major population-based data sources. Such groups of vulnerable children include those whose health may require special attention because of particular or multiple conditions of disadvantage, such as those in certain income cat- egories; those in certain racial or ethnic groups (such as American Indians or Alaska Natives); those who live in homes in which English is not the primary language spoken; those in residential or institutional care (such as foster care); those who are uninsured or underinsured; and those who reside in certain geographic areas, such as selected census tracts, rural environ- ments, or regions with low numbers of health care providers (underserved communities). Finally, in an ideal world, child and adolescent health measures would support analyses of the ways in which economic and social circumstances influence health status. Such analyses might include the relationships among children’s insurance status, their access to health providers, and their use of and the effectiveness of health care, as well as the relationship between child health status and family income, family stability and preservation, and children’s school readiness and educational achievement and attain- ment. The measures would also make it possible to examine relationships between the health status of children and adolescents and their educational performance, their social behaviors, and their future health status and pro- ductivity as adults. The remainder of this chapter examines the current status of child and adolescent health measures; measures of health care quality are discussed in Chapter 5. The first section takes a detailed look at existing measures, including their strengths and limitations. Issues of the timeliness, qual- ity, public transparency, and accessibility of currently available data on child and adolescent health are then addressed. Next, the chapter turns to the challenges of aggregating, synthesizing, and linking multiple sources of these data. This is followed by a review of efforts to make the data
OCR for page 95
95 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH more meaningful by linking population health indicators and public health interventions. EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH In preparing a review of existing measures of child and adolescent health, the committee identified seven priority areas for measurement, cur- rent related measures, and the existing sources that provide data on these measures. The priority areas are based on the committee’s collective judg- ment and emerged through careful deliberations, a thorough review of the literature, workshop presentations from a variety of engaged stakeholders and experts, and an extensive review of existing data sets. The committee considered the strengths and limitations of measures within each priority area, as well as the extent to which national and state-based data sources are available within each area. The seven priority areas are • c hildhood morbidity and mortality, • chronic disease conditions, • p reventable common health conditions (especially mental and be- havioral health and oral health), • functional status, • end-of-life conditions, • health disparities, and • social determinants of health. In addition, the committee considered the life-course approach, discussed in detail in Chapter 2, to be an overarching priority area that is integral to all seven areas listed above. The committee therefore contends that mea- surement should be informed by a life-course perspective and includes in this section a review of the limited number of existing measures and data collection efforts related to the life course. Using these priority areas as a starting point for examining the exist- ing array of measures and data collection efforts differs from previous approaches. For example, the IOM-NRC report Children’s Health, the Nation’s Wealth (2004) focuses on the specific measures of child health included in selected national surveys (e.g., up-to-date immunizations or nutrition adequacy). Instead, the approach used in this report enables those who are interested in a particular aspect of child and adolescent health (e.g., preventable common health conditions) to readily identify the most relevant currently available data sources. The sections that follow review child and adolescent health measures and data sources according to the seven priority areas, as well as the life-course approach; a more comprehensive review of the relevant data sets is included in Appendix D.
OCR for page 96
96 CHILD AND ADOLESCENT HEALTH Childhood Morbidity and Mortality A considerable amount of data related to child and adolescent morbid- ity and mortality is routinely collected and analyzed. Surveillance of injuries and fatalities among young people, for example, provides insight into one aspect of how children are doing and underscores how their epidemiol- ogy differs from that of adults. While unintentional injuries are a leading cause of death among Americans of all ages, they are the leading cause of death among children and adolescents aged 1−19 (Bernard et al., 2007) (see Box 4-1). Young children (under age 4) are especially vulnerable to life-threatening injuries (e.g., suffocation, drowning, and injuries related to motor vehicle crashes) (CDC, 2006). Three primary sources of data are used nationally to track morbidity and mortality: the National Vital Statistics System (NVSS), the Medical Expenditure Panel Survey (MEPS), and the Healthcare Cost and Utilization Project (HCUP). BOX 4-1 Leading Causes of Death Among Children and Adolescents Accidents* are by far the leading cause of death among children and adoles- cents. The top three causes of death by age group are listed below. Ages 0−1: • Developmental and genetic conditions present at birth • Sudden infant death syndrome • All conditions associated with prematurity and low birth weight Ages 1–4: • Accidents/injuries • Developmental and genetic conditions present at birth • Cancer Ages 5−14: • Accidents/injuries • Cancer • Homicide Ages 15−24: • Accidents/injuries • Homicide • Suicide * The preferred term for “accidents” is “unintentional injuries.” SOURCE: NIH, 2010b.
OCR for page 97
97 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH The NVSS is maintained by the National Center for Health Statistics (NCHS) within the Centers for Disease Control and Prevention (CDC). Federal reports frequently use data from the NVSS to monitor trends in child and adolescent mortality on a regional, national, and international basis. NVSS data are collected through ongoing reports from vital statistics officers in 50 states and the District of Columbia and reflect the cause of death that is recorded on individual death certificates, providing the basis for analyses of the leading causes of childhood morbidity and mortality. The data are organized by age and gender, as well as selected racial and ethnic groups. The NVSS relies on International Classification of Diseases (ICD) codes to describe health conditions, disorders, diseases, and injuries. For the most part, the ICD codes are organized by disease or injury catego- ries, such as different types of cancers or congenital conditions, infectious and parasitic diseases, endocrine conditions, mental disorders, disorders of pregnancy and childbirth, poisonings, drowning, and so forth. Hospitalization data for children and adolescents are collected through such data sources as the MEPS, as well as such syntheses of public−private data collection efforts as the HCUP. MEPS data are collected through a na- tionally representative survey of U.S. civilian households. The data provide information on the utilization and cost of health services, as well as on the cost, scope, and breadth of private health insurance held by and available to the U.S. population. HCUP data include a census of hospital discharge billing records collected from 40 states. The data provide information on reasons for hospitalization, length of hospital stays, procedures during hospitalization, and treatments received for specific conditions while in the hospital. As a part of HCUP, the Agency for Healthcare Quality and Research (AHRQ) developed a database specifically designed to allow in-depth stud- ies of children’s hospitalizations—the Kids’ Inpatient Database (KID). The KID is a stratified probability sample of pediatric discharges from 2,500– 4,000 community hospitals in the United States (defined as short-term, nonfederal general and specialty hospitals, excluding hospital units of other institutions). The purpose of KID data, which are drawn from an all-payer (Medicaid, private insurance, and uninsured) inpatient care database for children, is to identify, track, and analyze national trends in utilization, ac- cess, charges, quality, and outcomes for inpatient hospital services. Large claims-based data sets available from insurers and vendors also are commonly used in research on health care utilization and on preva- lence of disease. Examples include the Medstat Marketscan data set and the data sets of Blue Cross Blue Shield, Wellpoint/HealthCore, and Kaiser Permanente. Data collected by the HCUP and the KID reveal the most common reasons for admission to the hospital among children aged 17 and younger.
OCR for page 98
98 CHILD AND ADOLESCENT HEALTH The overwhelming majority—approximately 95 percent—of these admis- sions are for the birth of infants (Owens et al., 2003). Newborns, or chil- dren 30 days of age or less, account for approximately 4.8 million hospital stays or 73 percent of all childhood admissions (Elixhauser, 2008). Affective disorders, including depression and bipolar disorders, are the sixth most common reason for hospital admissions among children, accounting for 82,500 discharges. Adolescent pregnancy is one of the leading causes of hospitalization for females younger than 17. For adolescent boys, hospi- talization occurs primarily as a result of unintentional injuries (Owens et al., 2003). Strengths NVSS data provide a rigorous classification scheme for deaths associ- ated with an array of health conditions, including pregnancy, abortions, and various types of injuries that are common among children and adolescents. The data can be pooled and analyses conducted over multiple years by gender, race and ethnicity, and geographic location (state and county level) to highlight trends that may not be apparent within a single time period. The NVSS E-codes provide supplemental information about the cause of injury (such as motor vehicle crash or child maltreatment). The rigor of the data classification and the ongoing data collection support analyses of trends among racial and ethnic minority groups that are often difficult to detect in studies that rely on household surveys or other data sources. For example, one CDC study of fatal injuries among children by race and ethnicity (1999−2002) highlighted disproportionate rates of deaths due to motor vehicle injuries among American Indian/Alaska Native children, as well as higher rates of drowning deaths among black infants and American Indians/Alaska Natives aged 1−19 (Bernard et al., 2007). Linked death and birth records permit the examination of infant deaths by characteristics of the parents and can be used to compare the mortality experience of dif- ferent subpopulations (IOM, 1993). Linked records also provide insight into access to prenatal and delivery care and some outcomes of pregnancy (Marquis and Long, 2002; Schoendorf and Branum, 2006). Data collected through the MEPS and HCUP may be more accurate and reliable than survey data. For example, data obtained directly from providers, such as specific diagnoses and treatment, are less likely to be affected by recall bias than comparable data obtained from surveys based on self-reports (Cohen, 2004). Hospital discharge data can often be linked to other data sets, including data from the social services, criminal justice, education, housing, and other sectors (Schoenman et al., 2005). The KID’s large sample size enables analyses of both common and rare conditions. The database comprises more than 100 clinical and nonclinical
OCR for page 99
99 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH variables for each hospital stay, including primary and secondary diagnoses and procedures, admission and discharge status, patient demographics (e.g., gender, age, race, median income for ZIP code), expected payment source, total charges, length of stay, and hospital characteristics (e.g., ownership, size, teaching status). The KID contains clinical and resource use data included in a typical discharge abstract, but excludes data elements that could identify individuals directly or indirectly. Analyses of HCUP and KID data on rates of hospital admissions for specific conditions per popula- tion or rates of specific events per procedure can provide the hospital and reimbursement perspective on health care quality in terms of effectiveness and patient safety (Berdahl et al., 2010). Children can be identified by age in the Household Component of the MEPS, allowing most MEPS analyses to be performed for children. In 2001, a Child Health and Preventive Care section was added to the survey. It contains questions previously included in the 2000 Parent Administered Questionnaire, selected questions related to children that had been asked in previous years, and additional questions related to child preventive care. Limitations Morbidity and mortality data provide information for only the most severe health consequences, which involve a relatively small number of children and adolescents. Those who are concerned with children’s health status often want to know more than just the presence or absence of specific health problems in the general child population at a given point in time. They want to know the sequence of health conditions that may contribute to morbidity and mortality events, as well as the relationship between selected health conditions and certain social characteristics. They want to know whether children who have access to certain family resources, certain types of health care providers, or certain environmental and social condi- tions fare better than those who do not. And increasingly, they want to know whether children are on track to become healthy adults, especially those young people who display early signs of poor health conditions that are associated with adverse health outcomes and chronic disease in older populations. While NCHS can link vital statistics data with other data sources (in- cluding census data, Supplemental Nutrition Program for Women, Infants, and Children [WIC] program data, and hospital discharge data), NVSS data alone are limited in the information they can provide. For example, NVSS data do not capture fetal mortality experience by special populations (e.g., populations that are relatively small in number). Furthermore, challenges to data collection, including frequent item nonresponse, variation in state reporting requirements, and racial misclassification, may limit the overall
OCR for page 100
100 CHILD AND ADOLESCENT HEALTH quality and utility of NVSS data (Hoyert and Martin, 2002). The NVSS also does not collect information about family or other household charac- teristics (e.g., socioeconomic status), nor does it collect data on the types of health plans associated with selected health conditions or injuries. Hospital discharge data, of course, are limited in that they capture only those events that occur in a hospital. Moreover, the HCUP does not include data from all states, and less populous states are underrepresented. Further, the HCUP is not designed specifically for pediatric issues and does not allow for longi- tudinal studies of individuals. It is unclear whether the KID has the capacity to capture a representative sample of uncommon and rare diagnoses. Chronic Disease Conditions The number of children and youth in the United States identified as having chronic health conditions has increased considerably in the past four decades. Data from the 2009 National Health Interview Survey (NHIS), for example, indicate that 14 percent (more than 10 million) of children in the United States aged 17 and under have ever been diagnosed with asthma and that 10 percent (7.1 million) of children still have asthma. The 2009 survey also found that 9 percent (5 million) of children aged 3−17 had attention- deficit/hyperactivity disorder (ADHD) (Bloom et al., 2010). More than 12 million U.S. children meet the definition of children and youth with special health care needs—those at “increased risk for chronic physical, develop- mental, behavioral, or emotional conditions that require health and related services of a type or amount beyond that required of children generally” (McCormick et al., 2011; McPherson et al., 1998, p. 138). This group accounts for roughly 15−20 percent of the childhood population and for 80 percent of annual health care expenditures for all children (Newacheck et al., 1998b). Whether the increase in the number of children and adoles- cents with chronic health conditions is the result of environmental changes, better survival rates for once-fatal conditions, or increased access to care through Medicaid expansions and the Children’s Health Insurance Program (CHIP), it represents a significant trend (Van Cleave et al., 2010). The NHIS is conducted annually and collects data on health indicators, health care utilization and access (including current health insurance cover- age), and health-related behaviors for the U.S. civilian noninstitutionalized population. As a household survey, the NHIS collects data on all members of the household, including children, adolescents, and adults. Data collected through the NHIS are used to monitor trends in illness and disability and to track progress toward the achievement of national health objectives (Bloom et al., 2010). The National Survey of Children’s Health (NSCH), first introduced in 2003 and subsequently fielded in 2007, is one of the most comprehensive
OCR for page 101
101 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH surveys of child and adolescent health that offers national as well as state- level data (NCHS, 2010b). Data collected through the NSCH support analyses of physical, emotional, and behavioral child health indicators, as well as contextual factors. The next NSCH survey, planned for 2011, will expand the measurement of insurance adequacy beyond “having coverage” to include items regarding the actual providers and services covered by the child’s insurance policy, the costs of services not covered by the deductable, and the overall adequacy of benefits (Bethell and Newacheck, 2010). The NSCH is complemented by two other national surveys—the Na- tional Survey of Children with Special Health Care Needs (NS-CSHCN) and the National Survey of Early Child Health (NSECH). The NS-CSHCN was first conducted in 2001 and again in 2005−2006 to monitor states’ provision of services to children with special health care needs through federal programs, such as Title V and Supplemental Security Income (SSI) (Blumberg et al., 2003; van Dyck et al., 2002). The NS-CSHCN measures more than 100 indicators of children’s health and well-being for children enrolled in these programs in six key areas: health status, health care, school and activities, family and neighborhood, young children (aged 0–5), and school-aged children (aged 6−17). The NS-CSHCN was developed to measure the prevalence among children of both chronic conditions (e.g., asthma; attention-deficit disorder [ADD]/ADHD; depression, anxiety, or other emotional problems; mental retardation; and seizure disorders) and functional difficulties (e.g., respiratory problems, behavioral problems, chronic pain, and self-care), as well as services received and satisfaction with care (Blumberg et al., 2003; CAHMI, 2006; van Dyck et al., 2002). The NSECH is a nationally representative household survey of children aged 4−35 months that produces national and regional estimates. It was administered once, in 2000. Planning for a possible NSECH-II has been un- der way for several years, but no plan for its readministration has yet been developed. Survey questions include child developmental status, provision of recommended preventive services for which parents are valid reporters (e.g., anticipatory guidance, some screenings, and family-centered care), parenting behaviors and home safety, health insurance status, early child- hood program enrollment, and utilization of services (Halfon et al., 2002). The above three national surveys obtain national and state-based sam- ples that are weighted to represent the general population of noninstitu- tionalized children and adolescents. They all rely on a household survey platform known as the State and Local Area Integrated Telephone Survey (SLAITS), which is conducted by NCHS to support the design and sampling frame for the ongoing National Immunization Survey. The SLAITS operates by calling household telephone numbers at random to identify households with one or more children under 18. In each household, one child is ran- domly selected to be the subject of the interview.
OCR for page 124
124 CHILD AND ADOLESCENT HEALTH records, hospital discharge records) to follow their health outcomes and use of health care services (CDC, 2010c). Limitations The retrospective reporting of childhood experiences is a potential limitation of the ACE. Respondents may find it difficult to recall specific events. In cases in which childhood abuse has been documented, for example, adult respondents are likely to underestimate the actual occur- rence of the abuse upon follow-up (Femina et al., 1990; Williams, 1995). Another limitation relates to the sample included in the ACE. The major- ity of ACE participants are white (74.8 percent), middle-class adults, the overwhelming majority of whom have completed high school, attended college, or completed college and/or beyond (92.8 percent) (CDC, 2010a). These demographic characteristics limit the extent to which the findings of the study can be generalized. TIMELINESS, QUALITY, PUBLIC TRANSPARENCY, AND ACCESSIBILITY OF DATA ON CHILD AND ADOLESCENT HEALTH In its charge, the committee was asked to focus particular attention on the timeliness, quality, public transparency, and accessibility of data on child and adolescent health. Timeliness is a critical element in the assess- ment and development of measures, as more rapidly released public-use files provide a far more accurate picture of existing conditions than those released long after data collection (NRC, 2010). Public transparency de- pends on the timely availability and accessibility of quality data to reinforce accountability on the part of responsible agencies (Beal et al., 2004; IOM, 2001a). A number of online sources are designed to advance the timely and effective use of public data on children, youth, and families in the United States. Box 4-2 includes examples of accessible data sets across the seven priority areas for child and adolescent health that can be used by families, researchers, insurers, policy makers, and advocates to assess the health and mortality experiences of children and adolescents. These include public data sets, aggregations and syntheses of public data (see the next section), and sources that integrate public and private data. AGGREGATING, SYNTHESIZING, AND LINKING MULTIPLE DATA SOURCES Title V of the Social Security Act requires annual reporting of state performance and health outcome measurement data, fiscal data and num-
OCR for page 125
125 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH bers of clients served (individual, source, and service type), screening and treatment data, state priority needs, state Title V initiatives, maternal and child health (MCH) toll-free hotline data, and CSHCN service system data (MCHB, 2010). Although these data are posted in a timely fashion to the Title V Information System website, the data collected on child and adoles- cent health exist largely in individual silos and are not readily translatable to the seven priority areas discussed above. In the absence of population or administrative data sources that can link specific experiences or events to selected health behaviors in individual children, many researchers rely on linking selected data sources at the geographic level—for example, census tracts, counties, or states. Typically they link one of the individual-level data sources discussed above with an- other data source describing the social contexts of children and youth as proxy measures for adverse or supportive environments in a child’s census tract, county, or state. Such data include measures describing education, employment, income, and community crime trends for national or regional populations of children and youth. Box 4-3 provides examples of efforts to aggregate, synthesize, and link data from multiple sources. These include state, local, and national efforts using both publicly and privately collected data. Key sources of data for these efforts include the Current Population Survey (CPS), the Ameri- can Community Survey (ACS), the National Survey of American Families (NSAF), the National Center for Education Statistics (NCES) surveys, the NVSS, and the BRFSS, among others. Strengths Aggregating, synthesizing, and linking data from multiple data sources allows agencies and organizations to convey trends in child and adolescent health to policy makers and the general public. These efforts often generate easy-to-understand reports, fact books, and online tools. Limitations Unfortunately, linking multiple data sources cannot capture the dy- namics of child and adolescent health and does not provide insight into the interactions among various influences on child and adolescent health. The data sets are frequently based on cross-sectional data, a disadvantage for any effort to link multiple data sources. At present, moreover, financial barriers hinder the ability to access deidentified Medicaid files for purposes of cross-state quality measurement. As a result, current efforts to aggregate, synthesize, and link data result in something more akin to a mosaic than a snapshot of child and adolescent health, falling short of the goal of provid-
OCR for page 126
126 CHILD AND ADOLESCENT HEALTH BOX 4-2 Selected Online Sources of Data on Child and Adolescent Health • ehavioral Risk Factor Surveillance System (BRFSS) Interactive Data- B bases provide online access to the state-based system of health surveys that collects information on health risk behaviors, preventive health practices, and health care access primarily related to chronic disease and injury (http://www. cdc.gov/brfss/). • PONDER is a web-based query system created to access data collected C through Pregnancy Risk Assessment Monitoring System (PRAMS) surveys. Users have the ability to design their own analysis by choosing from an in- dexed list of available categorical variables. Descriptive statistics in the form of proportions are included in the resulting report and corresponding graph. CPONDER contains PRAMS data from 2000 through 2006 for state/year combinations that achieve at least a 70 percent response rate. CPONDER contains 2007 data for PRAMS state/year combinations that achieve at least a 65 percent response rate. As additional years of data are weighted, they will be added to the system (http://www.cdc.gov/prams/cponder.htm). • ATA2010 is an interactive database system developed by staff of the Division D of Health Promotion Statistics at the National Center for Health Statistics, and contains the most recent monitoring data for tracking Healthy People 2010. Data are included for all the objectives and subgroups identified in the Healthy People 2010: Objectives for Improving Health. DATA2010 contains primarily national data. However, state-based data are provided as available (http:// wonder.cdc.gov/data2010/). • he Data Resource Center for Child and Adolescent Health (DRC) pro- T vides online access to the survey data that allows users to compare state, regional, and nationwide results for every state and HRSA region as well as resources and personalized assistance for interpreting and reporting findings. DRC includes data from the National Survey of Children’s Health (NCHS) and the National Survey of Children with Special Health Care Needs (NS-CSHCN) (http://www.childhealthdata.org/content/Default.aspx). • CUPnet is a web-based interactive service for identifying, tracking, analyzing, H and comparing statistics on hospital care. HCUPnet was created with the inten- tion to make health care data available to the public. HCUPnet allows anyone to access aggregate statistics from these data sets to generate descriptive statistics on many topics of interest, including, for example, the percentage
OCR for page 127
127 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH of hospitalizations for children who are uninsured by state, trends in hospital admissions for specific conditions, quality indicators and information on the expenses of conditions treated in hospitals (http://hcupnet.ahrq.gov/). • H ealth Data Interactive presents tables with national health statistics for infants, children, adolescents, adults, and older adults. Tables can be custom- ized by age, gender, race/ethnicity, and geographic location to explore different trends and patterns (includes the following data sources: Current Population Survey [CPS], National Ambulatory Medical Care Survey [NAMCS], National Health and Nutrition Examination Survey [NHANES], National Health Care Survey [NHCS], National Health Interview Survey [NHIS], National Home and Hospice Care Survey [NHHCS], National Hospital Ambulatory Medical Care Survey (NHAMCS), National Hospital Discharge Survey [NHDS], National Vital Statistics System [NVSS] [mortality and natality], and population estimates) (http://www.cdc.gov/nchs/hdi.htm). • M EPSnet/HC is an interactive query tool that generates statistics of health care use, expenditures, sources of payment, and insurance coverage for the U.S. civilian noninstitutionalized population. However, none of the Child Health and Preventive Care section variables are available on MEPSnet/HC (http:// www.meps.ahrq.gov/mepsweb/data_stats/MEPSnetHC.jsp). • N ational Center for Health Statistics. Data files for the National Survey of CSHCN can be downloaded in SAS file format at no cost from the National Center for Health Statistics website (http://www.cshcndata.org). • N ational Immunization Survey Public Use Data Files are available for sta- tistical analysis or reporting purposes through the National Center for Health Statistics (http://www.cdc.gov/nis/data_files.htm). • W ISQARS™ (Web-based Injury Statistics Query and Reporting System) is an interactive database system that provides customized reports of injury-related data (http://www.cdc.gov/injury/wisqars/index.html). • Y outh Online is an online database allows users to analyze national, state, and local Youth Risk Behavior Surveillance System (YRBSS) data from 1991- 2009. Data from high school and middle school surveys are included. Users can filter and sort on the basis of race/ethnicity, sex, grade, or site, create customized tables and graphs, and perform statistical tests by site and health topic (http://apps.nccd.cdc.gov/youthonline/App/Default.aspx?SID=HS). NOTE: Descriptions are verbatim from source websites.
OCR for page 128
128 CHILD AND ADOLESCENT HEALTH BOX 4-3 Examples of Efforts to Aggregate, Synthesize, and Link Multiple Data Sources State and Local Governments/Health Departments California Report Card (Children Now) The Children’s Agenda (Montgomery County, Maryland) Children’s Score Card (Los Angeles County) Delaware Children’s Health Chartbook (Nemours) MassCHIP (Massachusetts Department of Public Health) North Carolina Child Health Report Card (Action for Children North Carolina, NC IOM) National America’s Children: Key National Indicators of Well-Being (Federal Interagency Forum on Child and Family Statistics) America’s Health Starts with Healthy Children: How Do States Compare? (Robert Wood Johnson Foundation) The Child and Youth Well-Being Index (The Foundation for Child Development) Child Health USA (Health Resources and Services Administration/ Maternal and Child Health Bureau) Child Trends DataBank (Child Trends) The Child Well-Being Index (The Foundation for Child Development) Indicators of Youth Health and Well-Being: Taking the Long View (Stagner and Zweigl, 2007) Key Indicators of Health and Safety: Infancy, Preschool, and Middle Childhood (Hogan and Msall, 2008) Kids Count (Annie E. Casey Foundation) Appendix A: Datasets for Measuring Children’s Health and Influences on Children’s Health, in Children’s Health, the Nation’s Wealth (IOM and NRC, 2004) Appendix B: Gaps Analysis of Measures of Children’s Health and Influences on Children’s Health in Select National Surveys, in Children’s Health, the Nation’s Wealth (IOM and NRC, 2004) Appendix C: Selected Indicators from National Children’s Data Syntheses, in Children’s Health, the Nation’s Wealth (IOM and NRC, 2004) ing a complete and accurate picture. Technology may make it possible to achieve this goal in the near future. Chapter 2 provides a brief overview of the implications of health information technology (HIT) for child and adolescent health. A more in-depth analysis of future implications of HIT for health and health care services is provided in Chapter 6.
OCR for page 129
129 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH EFFORTS TO MAKE DATA MEANINGFUL BY LINKING POPULATION HEALTH INDICATORS AND PUBLIC HEALTH INTERVENTIONS During the past three decades, efforts have been undertaken within public health and child advocacy centers to link population health data with national, state, and local initiatives designed to ameliorate those fac- tors that contribute to adverse health outcomes for children and youth. These efforts have emphasized identifying health conditions and behaviors that would benefit from public health interventions, as well as changes in social and economic settings, as opposed to medical treatments. Three such efforts are the Healthy People program, administered by CDC; County Health Rankings, developed within several states and published by The Robert Wood Johnson Foundation; and the Kids Count initiative, funded through the Annie E. Casey Foundation. The Healthy People 2010 and forthcoming Healthy People 2020 objec- tives provide a comprehensive agenda for nationwide health promotion and prevention of disease, disability, and premature death; they serve as a road map for improving the health of all Americans during the first decade of the 21st century. CDC relies extensively on health measures drawn from the NHIS and other data sources in the implementation of the Healthy People initiatives (HHS, 2000a). Healthy People 2010 includes 28 focus areas with 467 specific objec- tives. One of the 28 focus areas is maternal, infant, and child health, and 107 of the objectives pertain to adolescents and young adults. The two overarching goals of Healthy People, which are applicable across the life course, are to increase quality of life and years of healthy life and eliminate health disparities. A recent report on progress toward the Healthy People 2010 objectives describes mixed results for child and adolescent health. On the one hand, between 1996 and 2008, exposure of children to tobacco smoke at home and exposure to environmental tobacco smoke showed significant progress (reductions of 69.2 percent), and immunization of children aged 19–35 months increased by 10.9 percent. On the other hand, overweight in children and adolescents increased by 58.7 percent (Sondik et al., 2010). Efforts to finalize the Healthy People 2020 objectives have been under way since December 2010. Early indications point to a continued com- mitment to eliminating health disparities and a greater focus on the social determinants of health that have a disproportionate impact on specific racial/ethnic populations (Sondik et al., 2010). Two new overarching goals will be added: “promoting quality of life, healthy development, and healthy behaviors across life stages; and creating social and physical environments that promote good health” (Koh, 2010, p. 1656).
OCR for page 130
130 CHILD AND ADOLESCENT HEALTH The Robert Wood Johnson Foundation’s County Health Rankings ranks the overall health of every county in all 50 states. The rankings are based on a model of population health that includes health outcomes (based on equal weighting of length and quality of life) and health factors (weighted scores for health behaviors, clinical care, social and economic factors, and the physical environment) (see Figure 4-1) (Booske and UWPHI, 2010). The rankings are based on data from multiple sources, including • t he BRFSS; • t he NCHS; • t he National Center for Chronic Disease Prevention and Health Promotion (Division of Diabetes Translation); • t he National Center for Hepatitis, HIV, STD, and TB Prevention; • t he Environmental Protection Agency (EPA) Collaboration; • t he Health Resources and Services Administration; • t he CPS; • t he Federal Bureau of Investigation; • M edicare claims; and • t he National Center for Education Statistics. Bethell (2010) has identified four key questions to be considered in aligning population health indicators with efforts to improve the quality of health care services for children and youth: • S hould the emphasis be on leading causes of death and most com- mon reasons for using medical care or on the prevalence of ongo- ing health conditions (also described as the low-volume/high-cost versus high-volume/low-cost trade-off)? • S hould the population health measures be condition-specific (e.g., reflect the ICD categories), or should the broad-based, conse- quences-focused definition used in the survey of children with special health care needs (NS-CSHCN) be adopted? • W hat effort should be directed toward indicators of risk versus established conditions (e.g., overweight and obesity, or risks for developmental delay or substance use)? • S hould population health indicators aim to address categories of conditions (e.g., mental and behavioral health, oral health, injuries)? SUMMARY This chapter has reviewed the relative strengths and limitations of measures of the health of children and adolescents based on population
OCR for page 131
131 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH FIGURE 4-1 County Health Rankings model. SOURCE: Booske and UWPHI, 2010.
OCR for page 132
132 CHILD AND ADOLESCENT HEALTH health and administrative data sources. This review has highlighted the diversity and complexity of existing measures while calling attention to areas in which existing data systems are insufficient to address key topics of interest. For example: • A lack of standardization in the measurement of disparities in health limits the ability to identify, monitor, and address persistent health disparities among children and adolescents. • C urrent child health measures lack the capacity to capture impor- tant functional data and developmental stages; valid measures in these areas that have been tested across diverse populations do not yet exist. • M ost child and adolescent health data sets lack the capacity to support efforts to track the life-course implications of child health events, especially those that occur in early stages of development. The committee has identified seven priority areas for future measures that could provide relevant information on the health of children and ado- lescents for policy makers, service providers, and the general public and also inform quality improvement efforts within public and private health plans. The committee also has emphasized the importance of using a life-course approach, which may require changes to current public- and private-sector criteria and methods for the selection of existing and the development of new health quality measures. Indicators generated from data acquired with a life-course perspective in the seven priority areas should make it possible to examine specific conditions and issues of particular importance to vul- nerable and underserved children and adolescents, especially those served by Medicaid and CHIP programs. Such conditions and issues might include • g estational and perinatal issues that impact child health, such as prenatal care; • u nique neonatal issues, such as prematurity and low birth weight; • h ealth issues in the transition of those with chronic illnesses from adolescence to young adulthood (particularly in light of health re- form changes that include coverage of children under their parents’ health insurance until age 26); • c hronic childhood conditions that impact adult health, such as Down syndrome, cystic fibrosis, childhood cancer, and congenital heart defects; and • o pportunities presented by the NCS, which will follow subjects from preconception to age 21.
OCR for page 133
133 EXISTING MEASURES OF CHILD AND ADOLESCENT HEALTH Ideally, child and adolescent health quality measures should support analyses that can demonstrate how changes in funding levels for public insurance programs (such as Medicaid or CHIP) or changes in eligibility requirements, enrollment levels, or service procedures would affect child health outcomes, school achievement, and health care costs. Such measures should also be useful in assessing whether and how the organization and de- livery of health care achieve public goals of effectiveness, efficiency, safety, timeliness, equity, and patient satisfaction. Realizing these goals will require capacity for state-level analyses because Medicaid and CHIP are executed and managed at the state level, and there has historically been significant state-level variation in eligibility, coverage, and access to providers. Additional themes that deserve attention include the following: • t he distinction between low-incidence/high-cost conditions and those that reflect the most common child and adolescent health disorders; • s ignificant trends in child health, health care access and quality, and outcomes (e.g., immunization coverage rates); • i ndicators of resilience and protective factors/effects; and • c omorbidities (because of their potential multiplier effects). Finally, the seven priority areas, as well as a life-course perspective, should be used to direct analysis toward possible emerging threats to child health as a test of how comprehensive and useful this taxonomy can be in generating priority indicators for child and adolescent health.
OCR for page 134